Create an Iron Chef in Statistics Classes?
CAUSE Webinar
Rebekah Isaak
Laura Le
Laura Ziegler
& CATALST Team:
Andrew Zieffler
Joan Garfield
Robert delMas
Allan Rossman
Beth Chance
John Holcomb
George Cobb
Michelle Everson
June, 2011
DUE-0814433
Outline
✤
Introduction
✤
CATALST Research Foundations
✤
How We Create the Statistical Iron Chef
✤
Teaching Experiment
✤
Student Learning
✤
To Bring About Change…
Introduction
✤
Following a recipe step-by-step is to
“novice thinking” as understanding
affordances involved in truly cooking is
to “expert thinking”
CATALST Research Foundations
✤
Origins of CATALST
✤
George Cobb – new ideas about content
✤
Daniel Schwartz – “plowing the field”
✤
Tamara Moore – MEAs in other fields
CATALST Research Foundations
✤
Curricular materials based on research
in cognition and learning and instructional
design principles
✤
Materials expose students to the power of
statistics, real problems, and real, messy
data
✤
Radical changes in content and pedagogy:
No
t-Tests; randomization and re-sampling
approaches; MEAs
How We Create the Statistical
Iron Chef
✤
Model-Eliciting Activities (MEAs)
✤
Definition (from SERC website):
Model-eliciting activities (MEAs) are
activities that encourage students to
invent and test models. They are
posed as open-ended problems that
are designed to challenge students to
build models in order to solve complex,
real-world problems.
How We Create the Statistical
Iron Chef
✤
Model-Eliciting Activities (MEAs)
✤
Start each of three units with a
messy, real-world problem
✤
Example: iPod Shuffle MEA
✤
✤
Create rules to allow them to
judge whether or not the shuffle
feature on a particular iPod
appears to produce randomly
generated playlists.
End each unit with an “expert”
solution
http://serc.carleton.edu/sp/library/mea/what.html
How We Create the Statistical
Iron Chef
✤
Goals for the course:
✤
Immerse students in statistical
thinking
✤
Change the pedagogy and content
✤
Move to randomization/simulation
approach to inference
✤
Have students really “cook”
How We Create the Statistical
Iron Chef
✤
Unit 1: Models and Simulation
✤
Develop ideas of randomness and
modeling random chance
✤
Build an understanding of informal
inference that leads to an
introduction to formal inference
How We Create the Statistical
Iron Chef
✤
Unit 1: Models and Simulation
✤
Student Learning Goals:
✤
Understand the need to use simulation
to address questions involving
statistical inference.
✤
Develop an understanding of how we
simulate data to represent a random
process or model.
✤
Understand how to use the
results/outcomes generated by a
model to evaluate data observed in a
research study.
✤
Learn TinkerPlots
How We Create the Statistical
Iron Chef
How We Create the Statistical
Iron Chef
✤
Unit 2: Models for Comparing Groups
✤
Extend the concept of models and
formal inference by introducing
resampling methods
✤
Student Learning Goals
✤
Learn to model the variation due to
random assignment (i.e.,
Randomization Test) under the
assumption of no group differences
✤
Learn to model the variation due to
random sampling (i.e., Bootstrap
Test) under the assumption of no
group differences
How We Create the Statistical
Iron Chef
✤
Unit 3: Estimating Models Using Data
✤
Continue to use resampling methods
(i.e. bootstrap intervals) to develop
ideas of estimation
Teaching Experiment
✤
What is it?
✤
✤
They involve designing, teaching,
observing, and evaluating a
sequence of activities to help
students develop a particular
learning goal
2010/2011: Two-semester teaching
experiment (Year 3 of grant)
Preparation for the Teaching
Experiment

Reading, thinking, writing, adapting
MEAs

Planning and decisions about
sequence of course content, software
choice(s), etc.

Conversations and working sessions
with visiting scholars
Teaching Experiment: Semester 1
✤
Research Questions:
✤
How would students respond to the
demands of the course?
✤
What does it take to prepare
instructors to teach the course?
✤
How can we see evidence of the
students’ reasoning developing
throughout this course?
Teaching Experiment: Semester 1
✤
1 graduate student at UMN taught 1
section of undergraduate course (~30
students), while 2-3 graduate students
observed
✤
Unit 1 was written (and MEAs for Unit 2
and 3)
✤
Plans/Outline for Unit 2 and 3
✤
Plans for software (TinkerPlots, RTools, and R)
✤
Many weekly meetings to debrief and
plan
Ch-ch-ch-ch-Changes
✤
Team met in January to make changes
based on what was learned during the
semester (also met with 6 potential
implementers)
✤
Re-sequencing of some topics (e.g.,
bootstrap)
✤
Course readings added (content) and
removed (abstracts only)
✤
Assessments adapted as needed
✤
Group exams rather than individual
Teaching Experiment: Semester 2
✤
Research Questions:
✤
Is the revised sequence more coherent
and conceptually viable for students?
✤
How effective is the collaborative
teaching model in preparing instructors
for teaching the CATALST course?
✤
Can we take the experiences of these
instructors and use them to help create
lesson plans for future CATALST
teachers?
Teaching Experiment: Semester 2
✤
3 graduate students each taught a
section at U of M (~30 students each)
in active learning classrooms
✤
Also taught in 1 course at North
Carolina State University
✤
Many meetings (teaching team,
CATALST PIs, instructors, curriculum
writing, Herle Skype's into the meeting)
✤
Units 1 & 2 were written
✤
Plan/Outline for new Unit 3
Teaching Experiment: What We
Have Learned
✤
We can teach students to “cook”
✤
Based on interview and assessment
data, students seem to be thinking
statistically (even after only 6 class
periods!)
✤
We can change the content/pedagogy
of the introductory college course
✤
We can use software at this level that
is rooted in how students learn rather
than purely analytical
Student Learning: Positive Attitudes
Percent who selected Agree or Strongly Agree
COURSE EVALUATION ITEM
I feel that statistics offers valuable methods to analyze data to answer
important research questions.
I feel that as a result of taking this course, I can successfully use
statistics.
This course helped me understand statistical information I hear or read
about from the news media.
Learning to create models with TinkerPlots helped me learn to think
statistically.
Learning to use TinkerPlots was an important part of learning
statistics.
I think I am well-prepared for future classes that require an
understanding of statistics.
(N = 102)
95.0%
88.2%
86.3%
85.0%
81.4%
85.0%
Student Learning: Preliminary Results
✤
✤
Informal observations
✤
Different ways of answering the same
problem
✤
Small group discussions provide insight
into student thinking, particularly on
hard concepts
Student comments
✤
✤
✤
“I really didn’t anticipate enjoying a
stats class this much!”
“I would recommend this course to
anyone…I am very satisfied with this
course.”
“Really interesting way to learn
statistics!”
Challenges We are Working On
✤
Textbook/materials
✤
TinkerPlots™ scaffolding
✤
✤
✤
Get students to explore
Assessments
✤
Individual vs. cooperative
✤
Use of software on exams (not every student has a laptop)
✤
“Cheat” sheets
✤
Grading
Large courses
To Bring About Change…
✤
It takes a village
✤
It takes time
✤
It takes flexibility
Create an Iron Chef in Statistics
Classes?
YES!!!
http://catalystsumn.blogspot.com/
http://www.tc.umn.edu/~catalyst